> On Tue, Mar 10, 2026 at 12:43 PM Lorenzo Bianconi via dev <
> [email protected]> wrote:
> 
> > Introduce logical-flows incremental processing support for
> > logical_switch creation/deletion.
> >
> > Reported-at: https://issues.redhat.com/browse/FDP-2025
> > Signed-off-by: Lorenzo Bianconi <[email protected]>
> > ---
> >
> 
> Hi Lorenzo,
> 
> thank you for the patch, I have one question below.

Hi Ales,

thx for the review.

> 
> 
> >  northd/en-lflow.c     | 10 +++--
> >  northd/en-multicast.c |  3 +-
> >  northd/lflow-mgr.c    |  4 ++
> >  northd/northd.c       | 97 +++++++++++++++++++++++++++++++++++--------
> >  northd/northd.h       |  7 +++-
> >  tests/ovn-northd.at   | 11 +++--
> >  6 files changed, 105 insertions(+), 27 deletions(-)
> >

[...]

> > +                                    struct lflow_input *,
> > +                                    struct lflow_table *lflows);
> >  bool lflow_handle_northd_port_changes(struct ovsdb_idl_txn *ovnsb_txn,
> >                                        struct tracked_ovn_ports *,
> >                                        struct lflow_input *,
> > diff --git a/tests/ovn-northd.at b/tests/ovn-northd.at
> > index fd049d096..9d9a8cb84 100644
> > --- a/tests/ovn-northd.at
> > +++ b/tests/ovn-northd.at
> > @@ -2959,6 +2959,8 @@ OVN_FOR_EACH_NORTHD_NO_HV([
> >  AT_SETUP([check VXLAN mode disabling])
> >  ovn_start
> >
> > +OVS_CTL_TIMEOUT=120
> > +
> 
> 
> Why is increasing the timeout still necessary? I was under the
> impression that doing the unlink first and then sync in separate
> loop should solve the issue. But if that's not the case we need
> to definitely investigate and fix the slowness before considering
> this as a solution.

this is just a leftover of the previous implementation, I will remove it in v5.

Regards,
Lorenzo

> 
> 
> >  # Create a fake chassis with vxlan encap to implicitly enable VXLAN mode.
> >  check_uuid ovn-sbctl \
> >      --id=@e create encap chassis_name=hv1 ip="192.168.0.1" type="vxlan" \
> > @@ -2994,6 +2996,8 @@ OVN_FOR_EACH_NORTHD_NO_HV([
> >  AT_SETUP([check datapath tunnel ids exhaustion])
> >  ovn_start
> >
> > +OVS_CTL_TIMEOUT=120
> > +
> >  # Create a fake chassis with vxlan encap to lower MAX DP tunnel key to
> > 2^12
> >  check_uuid ovn-sbctl \
> >      --id=@e create encap chassis_name=hv1 ip="192.168.0.1" type="vxlan" \
> > @@ -3148,11 +3152,12 @@ check_row_count sb:logical_dp_group 1
> >  echo "$sw2_sb_uuid" >> sw_sb_uuids
> >
> >  AT_CHECK_UNQUOTED([ovn-sbctl --bare --columns _uuid,datapaths find
> > Logical_DP_Group dnl
> > -                    | grep -A1 $ls_dpg_uuid | tail -1 | tr ' ' '\n' |
> > sort], [0], [dnl
> > +                    | tail -1 | tr ' ' '\n' | sort], [0], [dnl
> >  $(cat sw_sb_uuids | sort)
> >  ])
> >
> >  dnl Add two routers and check that they are in their new separate group.
> > +ls_dpg_uuid=$(ovn-sbctl --bare --columns _uuid list Logical_DP_Group .)
> >  check ovn-nbctl --wait=sb lr-add lr0 -- lr-add lr1
> >  lr0_sb_uuid=$(fetch_column datapath_binding _uuid external_ids:name=lr0)
> >  lr1_sb_uuid=$(fetch_column datapath_binding _uuid external_ids:name=lr1)
> > @@ -14958,7 +14963,7 @@ check as northd ovn-appctl -t ovn-northd
> > inc-engine/clear-stats
> >  check ovn-nbctl --wait=sb ls-add sw0
> >  check_engine_stats northd norecompute compute
> >  check_engine_stats ls_stateful norecompute compute
> > -check_engine_stats lflow recompute nocompute
> > +check_engine_stats lflow norecompute compute
> >
> >  # For the below engine nodes, en_northd is input.  So check
> >  # their engine status.
> > @@ -14996,7 +15001,7 @@ check as northd ovn-appctl -t ovn-northd
> > inc-engine/clear-stats
> >  check ovn-nbctl --wait=sb ls-del sw0
> >  check_engine_stats northd norecompute compute
> >  check_engine_stats ls_stateful norecompute compute
> > -check_engine_stats lflow recompute nocompute
> > +check_engine_stats lflow norecompute compute
> >
> >  # For the below engine nodes, en_northd is input.  So check
> >  # their engine status.
> > --
> > 2.53.0
> >
> > _______________________________________________
> > dev mailing list
> > [email protected]
> > https://mail.openvswitch.org/mailman/listinfo/ovs-dev
> >
> >
> Regards,
> Ales
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to